Xinyu Li (Student)

Machine learning researcher with over 2 years of practical experience in deep learning, computer vision, and hardware-software co-design. Proficient in multiple computer languages such as C, Python, and Shell, with extensive experience in various projects including machine learning, computer vision, and semantic segmentation. Proficient in multiple languages, including Chinese, English, and Japanese. Capable of effective communication within a team and able to assume roles as a leader and coordinator. Possesses strong problem-solving skills, as well as a certain level of enthusiasm and ability for learning.

• Deep Learning Engineer, Traffic Sign Recognition Project, 01/2024, 12/2024, Curated ESRGAN-augmented GTSRB dataset (50 k images) and fine-tuned ResNet50 with mixed-precision on 4×A100 GPUs; test accuracy 99.2 %., Converted model to ONNX and deployed on NVIDIA Jetson Nano; sustained 30 FPS with 6 W power budget., Led 3-person team; established CI pipeline with GitHub Actions, reducing integration time by 40 %.
• Computer Vision Intern, Seal Segmentation Project, UNSW & Taronga Conservation, 01/2025, 12/2025, Built U-Net and SegFormer models to segment Antarctic fur seals in high-resolution satellite imagery; IoU 0.89 on unseen colonies., Implemented semi-automatic annotation tool using Label Studio & Python; boosted labeling throughput by 3 ×., Deployed trained model as Flask API on AWS EC2, enabling researchers to process 5 GB of imagery per hour.
• Course Project, Morse Code Encoder FSM, VHDL, 01/2024, 12/2024, Designed one-hot encoded FSM with shift-register and length-counter; verified timing and functional correctness in Quartus timing sim., Demonstrated on DE10-Lite board; latency deterministic within 1 clock.

• A method for eliminating blue light by adjusting the architecture of OLED products - Invention Patent, 2023
• A method for eliminating pixel dark spots based on OLED devices - Invention Patent, 2022
• Annual Outstanding Employee - BOE Group, 2023
• Outstanding Undergraduate Thesis Award, Dalian Maritime University, 2021